Pith. sign in

The PPLs can also guide the con- struction of CD-LM such as the datastore and retrieval modeling variations, to better fit the data of interest

[α2q2 + β2(1 − q2)] Indeed, any predictive probabilities can be computed as p(x∗ n:N |x∗ <n) =αnqn + βn(1 − qn) With this we can compute the perplexity (PPL) of any given text sequ

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
browse 1 citing papers

citation-role summary

background 1

citation-polarity summary

unclear 1

fields

cs.CL 1

years

2024 1

verdicts

CONDITIONAL 1

roles

background 1

polarities

unclear 1

representative citing papers

Chunk-Distilled Language Modeling

cs.CL · 2024-12-31 · conditional · novelty 6.0

CD-LM interleaves token-level generation with retrieval of multi-token chunks from a trie datastore, yielding exact sequence probabilities via a dynamic program and improving perplexity and inference speed without training.

citing papers explorer

Showing 1 of 1 citing paper.

  • Chunk-Distilled Language Modeling cs.CL · 2024-12-31 · conditional · none · ref 1

    CD-LM interleaves token-level generation with retrieval of multi-token chunks from a trie datastore, yielding exact sequence probabilities via a dynamic program and improving perplexity and inference speed without training.